Early intervention for schizophrenia is an increasingly influential movement in both research and clinical practice, and is based on the notion that treatment initiated during the prodromal phase of illness may prevent psychosis. However, little long-term prospective information is available to understand the nature, progression, or true outcome of the prodrome. Since not all individuals meeting established criteria will actually develop schizophrenia, there remains a primary need to identify those at true risk for later illness. This study is designed to prospectively characterize the developmental course of the prodrome and to determine which of its defining symptoms predict later emerging psychosis and schizophrenia. Further, this study will expand the range of outcomes beyond prediction of psychosis, to emphasize the severe functional impairments that characterize all phases of schizophrenia. Thus, the current request is for 5 years to substantially expand and enhance the study. Major goals of the proposed continuation include: 1) to validate a neurodevelopmental stage model of illness progression;2) to differentiate a "vulnerability core" of neurocognitive, behavioral and clinical risk factors that may serve as trait markers;3) to lengthen follow-up to 4 years, to more accurately assess outcome, with a major new focus on functional disability, and 4) to establish whether a composite index can be derived that will accurately predict which prodromal subjects will convert to psychosis. A total of 148 patients will be recruited over the course of Phase II and be combined with 152 participants already recruited as part of Phase I. This will generate a combined sample of 300 prodromal individuals-which will not only have considerable statistical power but also represents 1 of the largest prodromal samples in the literature. Based on our developmental strategy, the sample will be divided into 4 diagnostic stages, with 75 patients in each subgroup. In addition, a comparison sample of 100 healthy controls will be included. This design is expected to provide the prospective evidence base essential to develop effective, stage-specific interventions targeting a broad spectrum of outcomes. Prevention of psychosis remains a very long-term goal.